jlavarj

My one hang up about buying fist weapons is that they are MH or OH specific. A shaman in my guild had gotten the Merc Glad MH fist weapon, and then was power leveling blacksmithing for the drakefist/dragonmaw hammer when I pointed out that both of those were main hand specific weapons. It broke his heart. Taking the rippers leaves you without options should a better main hand come along, whereas with the axe and mace, you can just transfer the gladiator weapon into your offhand and not have to grind honor for another offhand.

 

Audrix

The only difference that would influence your picking of an axe or mace would be a racial. I believe orcs benefits from axes from what I have read. Other than that all 1h arena weapons are the same. The fist weapons as mention above would just lock it for MH use only.

 

Aspartem

Ive a question to the Addon "Enhancer". I dont know why, but i cant see the Enhancer-Points like, i think, Malan posted around Page 100.
Everytime i try to add the String from CrazyShaman, Enhancer says its an incorrect String. I'm lookin trhough this Thread and other Forums since 5-6 Hours - now i just ask for help^^

Could someone explain me, how this Addon works correctly?
Could i possibly be, that "Enhancer" dont work well with the German Version of WoW?

Greetz
Aspa

 

Patterns...

So I've searched the thread for some troubleshooting ideas for Yo!'s sim but couldn't find much. I've only been able to get it to run a sim once (out of *many* times messing with it). I read one post where someone said he clicked the "Theorycraft!" button 5 or 6 times until his CPU usage on task manager spiked to near 100%. This is exactly what happened the one time the sim ran for me -- since then though I haven't been able to get the CPU usage spike that would signify the sim was active. Which is to say, I don't believe it's a matter of the sim running and I just don't realize it; I think I can't get it to begin.

Anyone have any tips for working with the applet?

 

◊ Rob

If you have a Pentium 4 CPU with Hyperthreading, a Core Duo, or a Athlon 64 X2 your CPU usage will appear to spike only to 50%. Close WoW and any other applications that might use a lot of memory, even other Firefox windows -- I've noticed that the java applet is really unresponsive if you don't get it enough memory. Try using a low number of hours (100) and clicking the "Theorycraft!" button "firmly" until you see it locked in. It should come back relatively quickly. Then you can try using a higher number of hours for more reliable results.

 

Patterns...

Thanks Rob. So after another solid half-hour troubleshooting I discovered my silly mistake: gotta have values in all of the inputs below "Stats" on the Tracking Headquarters page. Evidently I'd been replacing the default 5 in the Haste% space with a null blank rather than a 0. Thought I'd share just in case anyone else has made/is making/will make the same error.

 

smokey

Do we have final EP values for T6 geared ?

 

rava

I don't see what purpose that serves. EP values are based on your stats and are used to determine if a piece of gear is an upgrade or not, yet you want the EP values of a shaman with no upgrades left. If I'm remembering right even with the values listed on the front page for entry level shaman the best items remain the best even with inflated/shifted EP values.

Full price for gum!? That dog won't hunt, Monsignor.

 

Lujaar

Yellow attacks have no crit cap because of the two-roll system, but white attacks converted from misses to hits by hit rating can't crit. I'm assuming that dodges converted to hits by expertise can't either.

So white crit cap is:

100% - 24% (glancing blows) - 28% (DW miss rate) - 5% (dodge) = 43%.

I can hit that twisting GoA with Kings and a feral druid. Even with no crit buffs whatsoever, double mongoose procs put me at just under 43%.

Ugh, guess I should look into getting an executioner enchant after all.

Edit: As Morelis and Yo pointed out, WoWWiki says I'm wrong about this. Crit cap is 43% + your hit.

 

Morelis

Really? If true that would be big news, but this is the first I've heard of it.

 

◊ Rob

No it wouldn't. That's the one-roll attack table in action, and it's probably the most basic assumption that this entire thread has been operating on.

 

Morelis

That would seem to conflict directly with the information here though that information is using old data, was this something that changed at the point the glancing rate was reduced? Just so I'm clear, you're saying that increasing hit has no bearing on the crit cap for white attacks?

edit: Just tried this with Yo's sim and using high crit rates shows that he is following the reasoning that hit does increase the crit cap for white attacks, if incorrect he should fix that.

 

• Scheme

No, what he's saying is that it wouldn't be big news, because it's a pretty integral part of melee-class theorycrafting.

 

Yo!

This link Weapon skill directly states
I fail to see
in Attack table
Here is dual-wielding example that shows that
Hit
Following this logic our crit cap is 55%-60% for most players..

 

◊ Rob

No, I'm saying that increasing that the statement "white attacks converted from misses to hits by hit rating can't crit" is the standard assumption. Normally, this is true. If you have 30% crit and 9% hit from talents and add 1% hit from rating, you will not gain any additional crits. This assumption has implicitly been made in every set of AEP valuations we've discussed, ranging from Disquette's original work to the numbers most people get from Yo's today.

This will only be false if you are crit capped. According to the article linked by Yo, the crit cap is 50.2%, before adding any hit rating. (Recall that glancing blows were dropped from occurring 40% of the time to ~20%.)

If you're properly specced you'll have at least 6% hit and probably 9% hit before any gear. So you're looking at 56%-59% crit cap before any gear. With Kings, LoTP, GoA, and double Mongoose rolling I'm just barely breaking 50% crit. Thus, "white attacks converted from misses to hit by hit rating can't crit". I would expect this to hold true, at least at the margin, for any shaman whose Armory profile you link me.


Edit:

Apparently, Lujaar was trying to argue that hit rating cannot increase the crit cap. The evidence I have seen suggests that this is incorrect (and the rest of this post makes the opposite assumption in calculating the crit cap). Morelis was right in correcting him.

I misunderstood what was being debated, and I took this to be another instance of "you need to hit to crit," which is a commonly held misconception we are confronted with weekly in this thread. Since the 2.1 glancing blows reduction, the crit cap has been a complete non-issue for enhancement shamans, and I haven't seen a post on it in months. I will leave both of my posts as-is in the hopes that we'll get one less "hit to crit" post.

To conclude:
* White melee attacks are resolved by a one-roll system in which your "paperdoll" critical strike percentage will occur as a percent of all swings, not of all hits.
* A theoretical "crit cap" does exist, but it is very high (52% using the numbers Lujaar gives for glancing blows, DW miss chance, dodge, and assuming 9% from talents).
* This crit cap increases as hit rating and expertise rating are obtained on gear. In practice, enhancement shamans will never reach this crit cap.

 

Yo!

+ Added Hand of Justice trinket as 1.33% on any attack for extra attack with same hand. Uses yellow attack hit chance. Adds as WF damage in report. Remember to add passive 20 ap directly
+ Added Elixir of Demonslaying
+ Changed Ashtongue Talisman proc rate to 50%
+ Changed Madness of Betrayer "real" (auto-attacks only) ppm to 1 to reflect upkeep of 40% as in game. 2.4 ppm that is used often for this trinket is total observed in game with yellow attacks, flurried white attacks, etc. and was incorreclty used by me
+ Changed DST "real" (auto-attacks only) ppm to 1 to reflect upkeep of 25% as in game (20 saec hidden cooldown)

The one proc that remains modelled not correctly is Dragonstrike.
Please post or direct me to the following data:
1) uptime that you observe in raids
2) uptime with auto-attacksonly
3) # of procs over 1 minute in raid
4) # of procs over 1 minute with auto-attacks only
5) any other info? (hidden cooldown, patch notes, etc.)

Though EP values of Elixir of Demonslaying and Mark of the Champion are trivial, I guess they are worth to be added to OP as they remain best options for specific fights.

 

Skeptikk

Thankyou for the updates Yo!

 

Malan

Hmm all the bosses in Hyjal count as demons right? Which ones in BT?

Shitting up every single thread on EJ since '06

 

Illundai

Mother Shahraz and Illidan, I'm guessing? I'm pretty sure Reliquary isn't a demon.

Btw, Rage Winterchill isn't a demon =P. All the rest is though.

Yes, Death Knight tanks are/were too good in 3.0.x. Having 38k HP, 76.81% Avoidance (Miss incl.) and not gimping your threat in the first tier of gear is broken. (Yes this is clickable)

 

Malan

Well I suppose at least 2 of the sunwell bosses are going to be demons too from what we know.

Shitting up every single thread on EJ since '06

 

• Anarkii

Supremus is a demon as well.

 

Lujaar

I was under the impression that converted hits had higher priority than crits, but I'm not going to argue with WoWWiki. Morelis and Yo, thanks for the information.

 

Visionnaire

Is there any reason to believe that the "offhand" Vengeful Gladiator weapon is any worse than the "one-hand" Vengeful Gladiator weapon for using in one's offhand?
(Considering they both have the EXACT same stats except for that label mentioned above)

 

spanko

Only the fact that its an offhand and you won't ever be able to use it your main hand.

 

Chopps

Thank you for all the wonderful information in this thread. I have questions regarding Yo! 's fantastic AEP Calculator I haven't been able to find here.

Am I correct in assuming the more "Hours of hack and slash" the more accurate my results will be? I know it sounds silly, but I'm getting drastically different values from running 20,000 hours from 100,000 hours. It would seem to make sense the more hours the better, but I wasn't sure if at some point too long of a test will mess something up. Throughout the thread, I've always seen people use 10-20K hours. I've followed the listed values on the first page for awhile, but I'd like to use my own values for evaluating upgrades.